Research on Expression Processing Methods of Children with Autism in Different Facial Feature Types

Facial expression recognition disorders were common in patients with Autism Spectrum Disorder (ASD), and were considered to be the main cause of their social disorders. In order to further understand the facial expression recognition disorders of children with ASD, this research adopted a 2 × 3 multi-factor experimental design and used eye-tracking technology to explore the expression processing of 3–7 years old ASD children and Typical Developing (TD) children with matching psychological development levels, under three different facial feature presentation types, including Low Spatial Frequency (LSF) facial expression stimulation, LSF and eye feature mixed facial expression stimulation, and LSF and mouth feature mixed facial expression stimulation. The results showed that ASD children’s visual scanning strategies for facial expressions were more disorganized than TD children. They were also weaker than TD children in processing facial expression structure information, but still had a “happiness advantage” similar to TD children, that was, the highest recognition rate of happiness expression. They were more accustomed to using local processing methods based on mouth features to process facial expression information, and their visual attention was more likely to be affected by the mouth rather than the eyes.